UZP (four registers)

Concatenate elements from four vectors

Concatenate every fourth element from each of the four source vectors and place them in the corresponding elements of the four destination vectors.

This instruction is unpredicated.

It has encodings from 2 classes: 8-bit to 64-bit elements and 128-bit element

8-bit to 64-bit elements
(FEAT_SME2)

313029282726252423222120191817161514131211109876543210
11000001size110110111000Zn00Zd10

UZP { <Zd1>.<T>-<Zd4>.<T> }, { <Zn1>.<T>-<Zn4>.<T> }

if !IsFeatureImplemented(FEAT_SME2) then UNDEFINED; if size == '11' && MaxImplementedSVL() < 256 then UNDEFINED; constant integer esize = 8 << UInt(size); constant integer n = UInt(Zn:'00'); constant integer d = UInt(Zd:'00');

128-bit element
(FEAT_SME2)

313029282726252423222120191817161514131211109876543210
1100000100110111111000Zn00Zd10

UZP { <Zd1>.Q-<Zd4>.Q }, { <Zn1>.Q-<Zn4>.Q }

if !IsFeatureImplemented(FEAT_SME2) then UNDEFINED; if MaxImplementedSVL() < 512 then UNDEFINED; constant integer esize = 128; constant integer n = UInt(Zn:'00'); constant integer d = UInt(Zd:'00');

Assembler Symbols

<Zd1>

Is the name of the first scalable vector register of the destination multi-vector group, encoded as "Zd" times 4.

<T>

Is the size specifier, encoded in size:

size <T>
00 B
01 H
10 S
11 D
<Zd4>

Is the name of the fourth scalable vector register of the destination multi-vector group, encoded as "Zd" times 4 plus 3.

<Zn1>

Is the name of the first scalable vector register of the source multi-vector group, encoded as "Zn" times 4.

<Zn4>

Is the name of the fourth scalable vector register of the source multi-vector group, encoded as "Zn" times 4 plus 3.

Operation

CheckStreamingSVEEnabled(); constant integer VL = CurrentVL; if VL < esize * 4 then UNDEFINED; constant integer quads = VL DIV (esize * 4); bits(VL) result0; bits(VL) result1; bits(VL) result2; bits(VL) result3; for r = 0 to 3 constant bits(VL) operand = Z[n+r, VL]; constant integer base = r * quads; for q = 0 to quads-1 Elem[result0, base+q, esize] = Elem[operand, 4*q+0, esize]; Elem[result1, base+q, esize] = Elem[operand, 4*q+1, esize]; Elem[result2, base+q, esize] = Elem[operand, 4*q+2, esize]; Elem[result3, base+q, esize] = Elem[operand, 4*q+3, esize]; Z[d+0, VL] = result0; Z[d+1, VL] = result1; Z[d+2, VL] = result2; Z[d+3, VL] = result3;

Operational information

If PSTATE.DIT is 1:


Internal version only: aarchmrs v2024-03_relA, pseudocode v2024-03_rel, sve v2024-03_rel ; Build timestamp: 2024-03-26T09:45

Copyright © 2010-2024 Arm Limited or its affiliates. All rights reserved. This document is Non-Confidential.